// RUN ON DATACENTER GPU

Run GemNet-T (OC20) on Datacenter GPU hardware

13 options ranked by estimated speed at Q4 / 8k context.

HardwareEst tok/sFits 8k?Max ctx~Price
NVIDIA B200 โ†’ ~6,541,353.4 yes 2672k โ€”
AMD Instinct MI325X โ†’ ~4,342,105.3 yes 3571k โ€”
NVIDIA H200 โ†’ ~3,924,812 yes 1956k โ€”
AMD Instinct MI300X 192GB โ†’ ~3,835,526.3 yes 2672k ~$15,000
NVIDIA H100 SXM5 80GB โ†’ ~2,739,191.7 yes 1100k ~$30,000
Intel Gaudi 3 โ†’ ~2,503,759.4 yes 1774k โ€”
NVIDIA A100 SXM4 80GB โ†’ ~1,667,227.4 yes 1100k ~$15,000
NVIDIA H100 PCIe 80GB โ†’ ~1,635,338.3 yes 1100k ~$25,000
NVIDIA A100 PCIe 80GB โ†’ ~1,582,189.8 yes 1100k ~$12,000
NVIDIA RTX PRO 6000 Blackwell โ†’ ~1,465,263.2 yes 1325k ~$8,500
NVIDIA RTX 6000 Ada 48GB โ†’ ~784,962.4 yes 651k ~$6,800
NVIDIA L40S 48GB โ†’ ~706,466.2 yes 651k ~$10,000
AMD Radeon Pro W7900 48GB โ†’ ~625,263.2 yes 651k ~$3,999

Estimate at Q4 (memory-bound). Click any option for its full per-hardware estimate and measured numbers.

FAQ

What Datacenter GPU hardware runs GemNet-T (OC20)?โ–ถ

On GenAIList, 13 Datacenter GPU options can run GemNet-T (OC20) at Q4 โ€” the fastest is the NVIDIA B200 at ~6,541,353.4 tok/s.

How fast is GemNet-T (OC20) on Datacenter GPU hardware?โ–ถ

Estimated generation speed ranges from ~625,263.2 to ~6,541,353.4 tok/s at Q4 (memory-bandwidth bound). Click any option for its full estimate and any measured numbers.